Search results

Dataloader segmentation fault

Status: Open.

#656 In karpathy/nanoGPT;

· mtodescato opened on Oct 31, 2025

MPS + float16: GradScaler / autocast behavior in train.py

Status: Open.

#654 In karpathy/nanoGPT;

· v659 opened on Oct 23, 2025

Feature request: semantic audit during training loop

Status: Open.

#653 In karpathy/nanoGPT;

· elly99-AI opened on Oct 22, 2025

Enable Private Security Reporting

Status: Open.

#646 In karpathy/nanoGPT;

· edoardottt opened on Oct 16, 2025

Loss Distribution at 400k and 600k steps - OWT - 124M model

Status: Open.

#644 In karpathy/nanoGPT;

· MarkoKarbevski opened on Oct 11, 2025

Use Nsight (nsys profile) to profile the training in a single multi-GPU node

Status: Open.

#642 In karpathy/nanoGPT;

· Minyuluck opened on Oct 1, 2025

Idea: Enhanced GPT with Mathematical Extensions (teaser)

Status: Open.

#637 In karpathy/nanoGPT;

· muzamil-ashiq opened on Sep 4, 2025

GUFF

Status: Open.

#636 In karpathy/nanoGPT;

· si-open opened on Sep 3, 2025

FutureWarning: <code>torch.cuda.amp.GradScaler(args...)</code> is deprecated.

Status: Open.

#635 In karpathy/nanoGPT;

· si-open opened on Sep 3, 2025

RuntimeError: Dataset scripts are no longer supported

Status: Open.

#633 In karpathy/nanoGPT;

· mlang opened on Aug 19, 2025

Sharing Learn-nanoGPT: An Educational NanoGPT Version with Hardware-Adaptive Training for Shakespeare Text

Status: Open.

#631 In karpathy/nanoGPT;

· niloydebbarma-code opened on Aug 16, 2025

NanoGPT for Verve

Status: Open.

#627 In karpathy/nanoGPT;

· CoderVanya opened on Aug 10, 2025